Study of area use and floor space occupation in office buildings using ML approach

Abstract

This thesis presents a general model to estimate the number of people at office spaces in the given time step. This project represents the description of the several approaches for similar problems, general description of statistical and machine learning models and applying those models for specific building. This work is also cover some suggested dashboards to keep track of space usage. The flowing models are applied to indoor air parameters: multi linear regression, support vector regression, and neural networks such as multi-layer perceptron and long-short term memory. The best performance was achieved by LSTM with 4 hidden layer and 16 number of neurons.